Create Your Own LUTs in Just a Few Clicks with fylm.ai – Tutorial

by Johnnie BehiriNovember 24^th, 2021

Filming in a flat picture profile can be a blessing as it opens up endless possibilities for creative grading yet at the same time can look really bad if done wrong. Each and every one of us has our own way of grading. Some will rely on free or paid downloaded LUTs, others will use third-party plug-ins to their NLE or grade from within their editing platform. All ways are legit, but I wanted to share with you what I’m currently using to successfully create my own LUTs in just a few clicks with fylm.ai. (For the complete tutorial, I recommend watching the video).

We all know the drill. As independent filmmakers, we film our material (many times in flat picture profile), edit, color grade, and distribute/upload our finished fabulous creation to wherever its final destination may be.

In the past, grading my footage was the step I felt most unsatisfied with. It simply never looked the way I wanted, or, I felt that my control over the final look has always been rather limited. At the same time, it was beyond my capacity to learn anything new, that required a “deep dive” and a steep learning curve although I acknowledge the need for it.

Around a year ago I was approached by Goran Ljubuncic, the man behind Lutify.me. He talked about his vision of developing a cloud base color grading application for “anyone”, from large studios to individuals like me. Of course, I love people with a vision, but at the same time, I’m a bit skeptical when I hear about a jack of all trade tool. A few lockdowns later and Goran invited me to a close beta group so I can experience firsthand his new ACES-compliant yet easy-to-use color grading application. And let me tell you, I was blown away. (ACES = Academy Color Encoding Specification.) Check out this video interview that I did with Goran a while ago about the new service.

In a nutshell, all that one has to do is import an image or video clip that was exported from the timeline into the fylm.ai app. Now, the app is a feature-rich color grading application, yet its strength lies in two separate working modes: “Magic” and “AI” (Artificial intelligence).

“Magic Mode” to create your own LUTs

The workflow in this mode is rather easy. After importing the image/video clip to the app, choose the flat picture profile you used while filming with your beloved camera and the right color temperature white balance (for example, S-Log3/daylight if you used a Sony camera your WB was set to daylight), then choose the type of your output monitor. (In my case, it is sRGB as I’m working on my laptop). From that point, just play with any of the dedicated twelve buttons. With these, one can adjust exposure, white balance, contrast, cine looks, print film emulation, skin tone, shadow tint, highlight tint, blue/green/yellow tone, and smart warmth.

The end result should be a self-made LUT that you can be proud of! One additional click and your newly created LUT can be downloaded as a .cube file that can be imported to any modern NLE.

“AI Mode”

The essence of this mode is to be able to give your video the look of any movie you desire. (Goran told me that A LOT of time went into “teaching the system” to do an accurate job to mimic specific looks). In the above video, I explain the process, but in general, after importing YOUR image or clip, you need to import the “other image” from the film you like and let fylm.ai analyze it and apply a matching color grade on YOUR image.

Needless to say that further color tweaks to the image you’ve created can be done on your NLE.

Final Thoughts

So far, this simple yet efficient color grading solution works very well for me and I simply encourage you guys to try this new cloud-based color grading app. Is there a discount offer for this service, you ask? Why yes! For a limited time only, get 50% off when subscribing to our MZed Pro Filmmaking Bundle (Black Friday special offer ending November 29th, 2021).

What are you guys using to color grade your videos? Did you already try using fylm.ai? Share your thoughts with us in the comment section below.